1. Cell Biology
  2. Developmental Biology and Stem Cells
Download icon

Ovulation in Drosophila is controlled by secretory cells of the female reproductive tract

  1. Jianjun Sun
  2. Allan C Spradling Is a corresponding author
  1. Carnegie Institution for Science, United States
  2. Howard Hughes Medical Institute, United States
Research Article
  • Cited 25
  • Views 1,610
  • Annotations
Cite as: eLife 2013;2:e00415 doi: 10.7554/eLife.00415


How oocytes are transferred into an oviduct with a receptive environment remains poorly known. We found that glands of the Drosophila female reproductive tract, spermathecae and/or parovaria, are required for ovulation and to promote sperm storage. Reducing total secretory cell number by interferring with Notch signaling during development blocked ovulation. Knocking down expression after adult eclosion of the nuclear hormone receptor Hr39, a master regulator of gland development, slowed ovulation and blocked sperm storage. However, ovulation (but not sperm storage) continued when only canonical protein secretion was compromised in adult glands. Our results imply that proteins secreted during adulthood by the canonical secretory pathway from female reproductive glands are needed to store sperm, while a non-canonical glandular secretion stimulates ovulation. Our results suggest that the reproductive tract signals to the ovary using glandular secretions, and that this pathway has been conserved during evolution.


eLife digest

Mammalian oviducts, or Fallopian tubes, convey egg cells from the ovaries to the uterus. Signalling between the ovary and oviduct, and secretory products produced throughout the reproductive tract, help to increase the likelihood of conception, minimise the loss of egg cells, and reduce the risk of ectopic pregnancy (in which an embryo implants outside the uterus). These processes may also influence the development of ovarian cancer, since Fallopian tube secretory cells were recently identified as the source of the most common and lethal subtype of epithelial ovarian cancer, high grade serous ovarian cancer.

Oviduct to ovary signalling is poorly understood in mammals. However, experiments using model organisms such as the fruit fly (Drosophila melanogaster) provide a potentially powerful approach to the problem, since many mechanisms in gametogenesis are conserved between species. In particular, secretions within the Drosophila female reproductive tract appear to boost reproductive success by interacting with sperm cells and seminal proteins, as in mammals. But whether these secretions reach the ovary and influence ovulation, or simply act on other aspects of reproduction such as mating, sperm storage, fertilisation or egg laying, remained unknown.

In this study, Sun and Spradling identified new genes controlling reproductive gland development and used this knowledge to elucidate secretory cell function. By mutating these genes, or the nuclear hormone receptor Hr39, they were able to reduce the total number of secretory cells that developed in the female reproductive tract, or to alter their function in adults. The ovaries of flies with abnormal secretory cell function contained as many egg cells as those of normal flies, but the mutant females laid fewer eggs. This indicates that secretory cells are required for at least one stage of reproduction.

By comparing ovulation rates in mutant and normal flies, Sun and Spradling showed that the secretory cells generate a product that is specifically required for ovulation, and that production depends on Hr39 activity. This Hr39-dependent secretion is a good candidate for a conserved signal between the reproductive tract and ovary because mouse Lrh-1, a mammalian gene closely related to Hr39, is expressed in oviduct secretory cells and is itself required for ovulation. The secretory cells were also found to produce protein secretions that are necessary for female flies to store sperm in the reproductive tract after mating.

By elucidating the roles played by female reproductive tract secretions, and demonstrating that they include a signal to the ovary that stimulates ovulation, the work of Sun and Spradling may lead to an increased understanding of ovarian cancer in humans.



The oviduct must interact extensively with the ovary to receive ovulated eggs in a manner that maximizes successful reproduction and minimizes egg loss and ectopic pregnancy. In humans, oviduct-ovary signaling also likely influence serous ovarian carcinoma development, a disease now thought to originate from the secretory epithelia of the distal oviduct (Fallopian tube) following an extended number of ovulatory cycles (Levanon et al., 2008; Bowtell, 2010; Kurman and Shih Ie, 2010; King et al., 2011; Berns and Bowtell, 2012). How the oviduct normally influences mammalian ovulation remains unclear, although a great deal has been learned about the hormonal control of ovulation within the ovary itself (Kim et al., 2009; Conti et al., 2012), and genes such as the nuclear hormone receptor Lrh-1 are known to be essential (Duggavathi et al., 2008). The growing realization that important aspects of gamete biology have been conserved during evolution suggests that insights into oviduct-ovary signaling may come from studies of model systems.

The Drosophila oviduct plays important roles during egg production that may involve communication with the ovary. The oviduct must be prepared to transport each oocyte released from the ovary to the uterus, to mediate its water uptake and eggshell crosslinking, and to position it for efficient fertilization (reviewed in Spradling, 1993). During each cycle of ovulation, just one of the many mature oocytes present in the two ovaries is released into an oviduct. Octopaminergic neurons innervating oviduct muscle and epithelia are needed for ovulation, probably to activate oviduct muscle contraction and to stimulate epithelial secretion by activating the Oamb octopamine receptor (Lee et al., 2003, 2009; Monastirioti, 2003). The steroid hormone ecdysone is produced in the adult ovary and is required to maintain egg production (Buszczak et al., 1999), although a specific role in ovulation has not been tested.

Glandular secretions from male reproductive tracts in both invertebrates and vertebrates facilitate reproduction at multiple steps (Bloch Qazi et al., 2003; Suarez, 2008; Heifetz and Rivlin, 2010; Holt and Fazeli, 2010; Ikawa et al., 2010; Jeong et al., 2010; Manier et al., 2010; Dunlap et al., 2011). Multiple proteins produced in the Drosophila male accessory glands are mixed with sperm upon ejaculation and transferred to the female reproductive tract where they mediate sperm storage, capacitation, and maternal reproductive behavior (Avila et al., 2011). For example, sex peptide (SP) increases egg laying and reduces female receptivity (Chen et al., 1988; Chapman et al., 2003; Liu and Kubli, 2003) by binding to a specific receptor, SPR, in three sets of fru+ppk+ sensory neurons in the female reproductive tract (Yapici et al., 2008; Hasemeyer et al., 2009; Yang et al., 2009). Ovulin, a protein transferred in male seminal fluid induces ovulation shortly after copulation (Herndon and Wolfner, 1995; Heifetz et al., 2005). Another transferred peptide, Acp36DE, facilitates sperm storage (Neubaum and Wolfner, 1999a; Avila and Wolfner, 2009). Ejaculate components produced in the mammalian testis, prostate, and epididymis also play important roles in reproduction (Suarez, 2008). For example, mammalian spermadhesins secreted from seminal vesicles mediate sperm attachment to the oviduct epithelia (Talevi and Gualtieri, 2010).

Female reproductive tract secretions also boost reproduction by interacting with transferred sperm and seminal proteins in many species (Holt and Fazeli, 2010; Jeong et al., 2010; Dunlap et al., 2011; Franco et al., 2011; Schnakenberg et al., 2011; Wolfner, 2011). Drosophila spermathecae and parovaria, the major exocrine glands of the female reproductive tract, are required for fertility and sperm storage (Anderson, 1945; Allen and Spradling, 2008; Schnakenberg et al., 2011). Whether Drosophila female secretory products regulate other aspects of reproduction remains poorly understood, however. Recently, reproductive secretory cell development in the spermathecae and parovaria was shown to follow a stereotyped cell lineage and to depend on the transcription factor Lozenge (Anderson, 1945) and Hr39 (Allen and Spradling, 2008; Sun and Spradling, 2012), a nuclear hormone receptor homologous to Lrh-1.

Here we used our new understanding of reproductive gland development to manipulate the number and activity of secretory cells in adult females. In addition to documenting a role for protein secretion in sperm storage, we show that adult Hr39 expression and a non-canonical secretion from the adult female reproductive glands are required for normal ovulation. Thus, ovulation in both Drosophila and mice depends on the homologous nuclear hormone receptors Hr39 and Lrh-1. Our results suggest that a conserved program of reproductive tract secretion mediates oviduct-ovary signaling and may be relevant to the origin of ovarian cancer.


Female reproductive glands are essential for ovulation

The overall function of female reproductive glands can be assessed by studying adult females bearing mutations in lz or Hr39 which disrupt their development (Anderson, 1945; Allen and Spradling, 2008). Mutants retain only rudimentary glands or lack them entirely and show defects in sperm storage and egg laying (Anderson, 1945; Allen and Spradling, 2008). All lz−/− females completely lack reproductive glands, while Hr39−/− females either lack glands (>90%) or retain a single defective spermathecae with very few secretory cells (Figure 1A–C). The ovaries in Hr39 and lz mutant females contain a full complement of mature oocytes, however, both lay significantly fewer eggs than controls (Figure 1D–E), indicating that secretory products are required for one or more steps downstream from oocyte completion, such as ovulation, mating, sperm storage, fertilization, or egg laying.

Female reproductive glands are essential for ovulation.

(A)–(C) DIC images of Oregon-R (A), Hr397154/Ly92 (B) and lz3/34 (C) mutant female lower reproductive tracts. Both spermathecae (yellow arrowheads) and parovaria (magenta arrows) are absent in the mutant animals. Bar graphs display the rate of egg laying (D and E), ovulation frequency (F and G), and copulation frequency (H and I) for the two mutant genotypes, and heterozygous controls. In all figures, the number of egg laying groups or mating pairs is shown in brackets. Error bars are SEM, or 95% confidence intervals. *p<0.01 (Fisher's exact test, or Student t-test).


It would be particularly interesting if reproductive glands were required for ovulation, since this might indicate that secretory products coordinate activities of the ovary and reproductive tract. Consequently, to distinguish whether ovulation was specifically affected, individual female flies were examined to determine if oocytes had left the ovary within 6 hr after mating to wild type males. In controls, about 50% of females initiated ovulation within this time interval, as indicated by the presence of an egg in either the oviduct, the uterus, or the food vial (Figure 1F,G). In contrast, none of the Hr39 mutant and only 2% of lz mutant females initiated ovulation. The failure of mutant animals to ovulate was not due to defects in mating, as both mutant and control females showed similar rates of copulation success (indicated by the presence of sperm in the female reproductive tract) (Figure 1H,I). Thus both Hr39 and lz are required for ovulation, at least initially.

lz and Hr39 are not required in reproductive tract neurons

Before determining which cells within the glands were needed for ovulation, we investigated whether the failure of lz and Hr39 mutant females to ovulate was due to a secondary requirement of these genes outside of the reproductive glands. For example, lz and Hr39 might function in the octopaminergic neurons that innervate oviduct muscle and stimulate oviduct epithelial cells prior to ovulation (Lee et al., 2003), or in the fru+ppk+ sensory neurons of the female reproductive tract. However, lz expression could not be detected in the octopaminergic neurons innervating the oviduct nor in the oviduct muscle or epithelial cells (Figure 2A). Lineage tracing also showed that oviduct cells are not derived from lz+ cells (Sun and Spradling, 2012). Yet blocking the proliferation of lz-expressing cells during pupal development was sufficient to disturb ovulation (Table 1).

lz and Hr39 are not required in reproductive tract neurons.

(A) lz expression (lzGal4 driving UAS-mCD8::GFP) in control female reproductive tract. Spermathecae (yellow arrowheads); parovaria (magenta arrowheads). Ov: Ovuduct; SR: Seminal receptacle; Ut: Uterus. Two sets of lz+ sensory neurons are illustrated at higher magnification in (A1 and A2). (B) lz expression in female reproductive tract expressing lzGal4>UAScycA (lz>cycARNAi). lz+ sensory neurons are not affected (B1 and B2). (C) Egg production is not affected by expressing Hr39-RNAi in ppk+ neurons of the reproductive tract. (D) Ectopic expression of mSP in fru+ reproductive tract neurons reduces virgin female copulation rate, even when neurons are mutant for Hr39. (E) Ectopic mSP in fru+ neurons is sufficient to induce egg laying in control virgin females but not in Hr39−/− females even in the presence of males. * indicates p<0.01 and NS indicates p>0.05.

Table 1

The effect of altering secretory cell (SC) number in female reproductive glands on egg laying, ovulation, copulation, and sperm storage in spermathecae

GenotypeFemale glandsEgg laying in 2 daysOvulation in 6 hrCopulation in 6 hrSperm storage in 6 hr
NSC/Female (Avg. ± SD)NEggs/female/Day (Avg. ± SEM)NOvulation (%)NCopulation (%)NSpermathecae with sperm (%)
lzGal410197.0 ± 18.04538.9 ± 3.93076.71889.0
lz>cycARNAi362.0 ± 2.6*251.0 ± 1.0*303.3*2356.5
lz>hr39RNAi1510.4 ± 7.4*255.2 ± 0.8*3020*1369.2
lz>hntRNAi12311.2 ± 8.4*458.0 ± 1.9*306.7*27100.0
lz>hntRNAi22239.4 ± 12.1*2516.5 ± 1.5
lz>PsnDN251.8 ± 1.9*252.0 ± 1.8*254*25100.0
lz>Su(H)DN160.9 ± 1.0*152.5 ± 2.0*
dpr5Gal45192.0 ± 15.44043.0 ± 4.53164.525100.05098.0
dpr5>cycARNAi259.3 ± 3.7*253.3 ± 1.5*358.6*3591.46515.4*
dpr5>hr39RNAi10191.3 ± 15.31542.1 ± 2.62445.82190.53783.8
dpr5>hntRNAi11817.1 ± 6.3*508.4 ± 1.8*342.9*3494.1649.4*
dpr5>hntRNAi21399.5 ± 20.6*2550.4 ± 2.82479.22495.84981.6
dpr5>NRNAi259.0 ± 3.2*250.9 ± 0.6*251*2391.3424.8*
dpr5>PsnDN1683.9 ± 11.5*2537.4 ± 7.52646.226100.05288.5
dpr5>Su(H)DN2016.1 ± 4.5*2526.2 ± 1.7147.1*1492.92615.4*
  1. *

    p<0.001. T-test was used for secretory cell number and egg laying. Fisher's exact test was used for ovulation, copulation, and sperm storage.

  2. p<0.01.

Further experiments argued against a requirement of lz and Hr39 in the fru+ppk+ sensory neurons for the female postmating behaviors elicited by the SP/SPR signaling pathway (Hasemeyer et al., 2009; Yang et al., 2009). lz was found to be expressed in a subset of these neurons near the oviduct-uterus junction, but fru+ppk+ neuron number was not affected by knocking down cycA or Hr39 using the lzGal4 driver (Figure 2A,B). Likewise, Hr39 does not function in these neurons because no defects were observed in egg laying when Hr39 levels were reduced by driving Hr39RNAi expression using ppkGal4 (Figure 2C). Furthermore, ectopically expressing a membrane-attached form of sex peptide (mSP) in fru+ neurons blocked virgin female receptivity even when carried out in an Hr39 mutant background (Figure 2D) (Hasemeyer et al., 2009; Yang et al., 2009), indicating that Hr39 mutant females have intact neural circuitry for post-mating behavior. Yet these same females still did not lay eggs (Figure 2E). Thus, the disruption in ovulation observed in lz and Hr39 mutant females is unlikely to be caused by altered SP/SPR signaling or to other neural defects within the reproductive tract.

Notch signaling and Hindsight are required to form reproductive gland secretory cells

In order to analyze which cells within the reproductive glands are required for ovulation, we developed methods for perturbing gland development more precisely than is possible using lz and Hr39 mutations. Several previous observations during studies of pupal spermathecal and parovarial development (Sun and Spradling, 2012) revealed a likely role for Notch signaling in their developing secretory cells (SCs). The stereotyped divisions of secretory unity precursor cells (SUPs) resemble the Notch-requiring divisions during peripheral nervous system development (Lai and Orgogozo, 2004; Sun and Spradling, 2012). Consistent with this idea, we discovered that a Notch signaling reporter is dynamically expressed in this lineage (Figure 3A–D, green). Hindsight (Hnt), a transcription factor that acts downstream from Notch during ovarian follicle cell development (Sun and Deng, 2007), was also expressed in developing and adult SCs but not epithelial cells (Figure 3B–D). Within the SUP lineage, secretory cells displayed the highest level of Notch activity and Hnt expression (Figure 3D,D').

Notch signaling and Hindsight are required to form reproductive gland secretory cells.

(A) The cell lineage underlying secretory cell development (Sun and Spradling, 2012). Notch signaling activity (green); Hnt expression (red). Ac: Apical cell; Bc: Basal cell; LEP: Lumen epithelial precursor; Sc: Secretory cell; SUP: Secretory unit precursor. (B)–(D) Notch activity (green) and Hnt (red) in spermathecae of adults (B), 26 hr pupae (APF) (C), and 48 hr APF (D). (D') shows the boxed region from (D). Yellow arrowhead: Epithelial cell. (E)–(G) Adult spermathecae from females expressing lz>Psn[DN] (E) or lz>hntRNAi (FG) during gland development. lz (green) marks epithelial cells; Hnt (red) marks secretory cells.


We extensively documented that Notch signaling and Hnt function during gland development using knockdown experiments (Table 1). Expression of NotchRNAi driven by lzGal4 causes pupal lethality, however, flies in which Notch signaling is disrupted using dominant negative forms of the pathway components Psn or Su(H) survive to adulthood. When we examined the reproductive glands in females of these genotypes, no secretory cells were observed, the gland lumen was collapsed and the duct was malformed (Figure 3E and Table 1). Depletion of Hnt with two different hntRNAi lines driven by lz-Gal4 almost completely blocked secretory cell formation, while the gland lumen and duct developed normally (Figure 3F,G and Table 1). It would be worthwhile to further investigate the roles Notch signaling plays during specific steps in the secretory cell lineage. Because, these differential cell divisions (Figure 3A) probably resemble those extensively characterized during peripheral nervous system development, we initially focused on using this new information to generate glands containing reduced numbers of secretory cells, without disturbing the epithelial portion of the gland.

Female reproductive tract secretory cells regulate ovulation

Adult females whose reproductive glands are deficient in secretory cells were generated by knocking down hnt expression during pupal development using a lzGal4 driver (Figure 3F,G), and tested for their ability to ovulate and lay eggs. SC-deficient females showed strong ovulation defects and laid significantly fewer eggs than controls (Table 1), indicating that secretory cells per se are required for ovulation. Females whose reproductive glands lack secretory cells were independently generated by expressing dominant negative (DN) forms of Psn (Figure 3E) or Su(H), and these females also had greatly reduced ovulation and laid few eggs (Table 1).

To further limit possible secondary defects present in animals that develop with reduced numbers of secretory cells, we searched for Gal4 drivers expressed specifically in female reproductive gland precursors among the Janelia Gal4 collection (Pfeiffer et al., 2008). From approximately 1000 lines screened, one Gal4 driver, 51B02 (termed dpr5Gal4) is specifically expressed in developing but not in mature secretory cells, nor in other reproductive tract or ovarian tissue (Figure 4—figure supplement 1). Using dpr5Gal4 to drive a lineage marker confirmed its specificity for the secretory lineage of the reproductive tract and its absence in sensory neurons (Figure 4—figure supplement 2). Females with different numbers of secretory cells were generated by knocking down cycA, Hr39, hnt, N, Psn, or Su(H) expression with dpr5Gal4 (Table 1). Regardless of which gene was targeted, the ability of these females to ovulate and to lay eggs depended on the number of secretory cells in their reproductive glands (Figure 4A,B). Copulation was not affected (Figure 4C). These results demonstrate that one or more products produced in the secretory cells of the reproductive tract are required for adult Drosophila females to ovulate and lay eggs.

Figure 4 with 2 supplements see all
Female reproductive tract secretory cells mediate ovulation and sperm storage.

(A)–(C) Relationship between secretory cell (SC) number and egg laying rate (A); percent ovulation (B); or percent copulation (C). Pooled data from genotypes in Table 1. Female reproductive tracts (D and E) and spermathecae (yellow circles in D and E; shown at higher magnification: D' and E') from normal females (dpr5Gal4 alone) (D) or females lacking SCs (dpr5Gal4>hntRNAi) (E) 6 hr after mating to males whose sperm nuclei are marked with protB-GFP (green). Seminal receptacle (white arrow). (F) Relationship between secretory cell number and the percentage of spermathecae with >5 sperm. Pooled data from genotypes in Table 1.


Female reproductive tract secretory cells are needed for sperm storage

Studies of animals lacking reproductive secretory organs (Anderson, 1945; Allen and Spradling, 2008) and of adults whose spermathecal secretory cells were partially ablated (Schnakenberg et al., 2011) have strongly argued that reproductive secretory cells produce products involved in sperm storage. We examined spermathecae for the presence of stored sperm in females generated as described above using dpr5Gal4 that differ in secretory cell number. Within 6 hr after copulation, most wild type females had finished transferring sperm from the uterus to the storage organs (seminal receptacle and spermathecae) (Figure 4D and Table 1) (Manier et al., 2010). In contrast, less than 20% of females with a severe deficit of secretory cells (e.g., dpr5Gal4>hntRNAi1) had sperm inside the spermathaecal lumen at this time (Figure 4E and Table 1). Even among rare spermathecae that contained sperm inside the lumen, the number stored was much less than in controls. The absence of stored sperm is unlikely to be due to a physical block to the spermathecae, since sperm were found in the spermathecal duct and were stored in the seminal receptacle (Figure 4E). Our experiments showed that a minimum of about 80 secretory cells are needed for females to store a normal number of sperm (Figure 4F), indicating that a quantitative requirement exists for the products of these cells.

Reproductive gland secretion and Hr39 action in adults are required for sperm storage

To further investigate the role of reproductive gland cells in adult female fertility, we sought to disrupt the activity of these cells during adulthood in glands that had developed normally. We shifted conditional mutations to the restrictive temperature only after eclosion (Figure 5A) and also used the Gal4 line syt12Gal4, which is expressed in mature secretory cells (Figure 5B), but not in the rest of the reproductive tract and ovary, to limit manipulations to adult secretory cells. The process of canonical protein secretion via the ER/Golgi/plasma membrane pathway was disrupted by expressing a dominant negative temperature sensitive allele of dynamin (shits) (Yang et al., 2009), or by knocking down the betaCOP or sec23 genes using RNAi controlled by the temperature sensitive Gal80 repressor (Lee et al., 2004; Bard et al., 2006; Aikin et al., 2012). When syt12Gal4 UAS-shits adults were raised to the non-permissive temperature at eclosion, membrane trafficking in secretory cells was rapidly disrupted as expected (Figure 5—figure supplement 1). After 6 days at the non-permissive temperature, these females showed severe defects in sperm storage within the spermathecal lumen (Figure 5B–D), and those sperm that were stored exhibited an abnormal morphology characterized by a twisted sperm head (Figure 5E–F). Despite this, the females contained many sperm within the seminal receptacle (Figure 5C) and laid a near normal number of eggs (Figure 5G). Even stronger reductions in the number of sperm within the spermathecae were observed when betaCOP or sec23 were knocked down in adults following the temperature shift to inactivate Gal80 (Figure 5D).

Figure 5 with 1 supplement see all
Canonical protein secretion from glandular secretory cells is required for sperm storage but not for ovulation.

(A) Experimental scheme for testing adult secretory cell function using temperature sensitive shits or GAL80ts. (B) and (C) Dynamin (Shi) is required for sperm storage. Female reproductive tract of syt12Gal4 control (B) or syt12Gal4 driving shits expression (C) 6 hr after mating to protB-GFP males at 29°C. syt12Gal4 expression is restricted to secretory cells as showed by UAS-RFP (red). (B') and (C'): Higher magnification of boxed spermathecae; seminal receptacle contain sperm (white arrows). (D) Sperm content of spermathecae (three classes) is reduced in flies with indicated genotype (x axis) at 29°C. Bracket: Number analyzed. *p<0.01 (chi-square test). (E) and (F) Abnormal morphology of spermathecal sperm in shits females at 29°C (F) compared to control (E). Egg laying rate (G) and ovulation time (from Table 2) (H) in flies with the indicated genotypes (x axes). *p<0.05 (Students t-test or Fisher's exact test). (I) Secretory cells use distinct secretory pathways to control sperm storage and ovulation.


This experimental paradigm also revealed an ongoing requirement for Hr39 in adults. When Hr39 function was knocked down in adult secretory cells under the control of Gal80ts (Figure 5A), severe reductions in the number of sperm stored in the spermathecae were also observed (Figure 5D). Those sperm that were present showed the same morphological defects seen in animals where canonical secretion had been reduced.

Non-canonical protein secretion in reproductive secretory cells is required in adults for normal ovulation

The effects on ovulation of knocking down Hr39 expression or disrupting canonical protein secretion were particularly interesting. We modified our ovulation assay so that it could be applied not only to the initial oocytes, but to ongoing ovulation throughout several days of mature adulthood (see ‘Materials and methods'). By determining the total number of eggs laid as well as the steady-state fraction of females that contained an egg in the uterus, we could calculate the average time oocytes spend during ovulation and within the uterus (Table 2).

Table 2

The effect of disrupting protein secretion or Hr39 expression during adulthood on the rate of egg laying and uterine egg content

GenotypeEgg laying in 2 days*Egg distribution in 6 hrEgg laying time (min)
NEggs/female/dayNUterus with egg (%)Total timeOvulation timeUterus time
syt12Gal42577.3 ± 2.32842.9 ± 18.317.1 ± 0.59.8 ± 3.17.3 ± 3.1
syt12>shits2572.2 ± 2.43268.8 ± 16.118.3 ± 0.65.7 ± 2.912.6 ± 3.0
syt12Gal42565.4 ± 2.32962.1 ± 17.720.2 ± 0.77.6 ± 3.612.5 ± 3.6
syt12>βCOPRNAi2553.8 ± 6.32885.7 ± 13.024.5 ± 2.93.5 ± 3.221.0 ± 4.0
syt12>sec23RNAi2551.5 ± 6.22986.2 ± 12.625.7 ± 3.13.5 ± 3.322.1 ± 4.2
syt12>Hr39RNAi2535.8 ± 2.23026.7 ± 15.836.9 ± 2.327 ± 6.19.8 ± 5.9
  1. *

    1 day = 22 hr at 29°C.

  2. p<0.05. All data are mean ± 95% confidence interval. T-test was used for egg laying, while Fisher's exact test was used for egg distribution.

Interestingly, females with ectopic adult expression of shits in secretory cells were not defective in egg laying or ovulation (Figure 5G,H) despite their compromised ability to store sperm in the spermathecae. In particular, the time required per ovulation event was not increased compared to control (Figure 5H). Females in which canonical secretion was disrupted by knocking down betaCOP or sec23 also laid eggs and ovulated similarly to controls (Figure 5G,H). Nonetheless, the secretory cell requirement that we had previously documented for the early ovulation was confirmed when we knocked down Hr39 in adult secretory cells. After six days at the non-permissive temperature, these animals showed a significantly lower rate of egg laying, about half of normal (Figure 5G) and ovulated much more slowly than controls (Figure 5H), requiring an average of 27 ± 6.1 min per egg compared to 7.6 ± 3.6 min in controls (Table 2). Our results show that at least two types of secretory cell products are released from reproductive gland secretory cells using different mechanisms. Products produced by the canonical protein secretory pathway are required to attract and store sperm in the spermathecae and to maintain their normal morphology. In addition, products needed to achieve a normal rate of ovulation require the function of Hr39 in secretory cells, but do not utilize the canonical secretory pathway.


Drosophila female reproductive tract secretions are required to attract and store sperm

Our experiments extend previous knowledge about the role reproductive tract secretions play in storing sperm. Sperm storage in the female reproductive tract is a general phenomenon in the animal kingdom including humans and insects (Neubaum and Wolfner, 1999b). In mammals, carbohydrate-dependent binding of sperm to the oviduct epithelia is important in order to form a sperm reservoir (Talevi and Gualtieri, 2010). In the absence of glands and hence of secretions, Drosophila sperm are still stored in the seminal receptacle, but they are poorly motile and fertility is low (Anderson, 1945; Allen and Spradling, 2008). The same outcome is observed when spermathecal secretory cells are partially ablated in adults prior to mating (Schnakenberg et al., 2011).

The work reported here allows several additional conclusions. First, the initial attraction of sperm to the spermathecae within 6 hr of mating requires a minimum amount of secretion from the secretory cells (SCs). Females with fewer than 25 SCs rarely contain sperm. In contrast, females with 80 or more SCs show the same high frequency of sperm in their spermathecae as wild type. The secreted attractive factor might interact with the accessory gland protein Acp36DE to facilitate uterine contraction (Avila and Wolfner, 2009), or act directly on sperm to regulate flagellar function (Kottgen et al., 2011; Yang et al., 2011). Second, the fact that sperm still move to the seminal receptacle in the absence of SCs shows that different mechanisms are involved in transporting sperm to the two different storage organs. Third, we found that female reproductive tract secretions are required to maintain sperm structurally. In the absence of secretory cells, sperm are not attracted to the spermathecae and those in the seminal receptacle aggregate and are difficult to individually assess (Anderson, 1945; Allen and Spradling, 2008). However, when protein secretion in SCs is disrupted using shits, sperm that do make it to the spermathecae exhibit distinctive morphological abnormalities.

Finally, we documented that secretory cells competent to carry out canonical protein secretion and expressing Hr39 are required to store sperm during adulthood. When protein secretion was disrupted after eclosion, sperm storage in the spermathecae was drastically compromised. Since we did not mate these females until the day they were tested, our experiments show that any initial accumulation of secretory products during pupal development turns over or is insufficient to store new sperm. The requirement for new secretion from the reproductive glands is consistent with previous studies showing that some proteins in these glands are induced by mating (Mack et al. 2006). The fact that Hr39 is required confirms that this gene, which is known to play a prominent role during reproductive gland development and to be expressed in adult secretory cells (Allen and Spradling, 2008), does play a key functional role in the adult gland. Many but not all spermathecal protein mRNAs, including many that are likely to be involved in sperm maintenance, are greatly reduced in an Hr39 mutant that retains spermathecae (Allen and Spradling, 2008).

Drosophila female reproductive tract secretions are required for ovulation

A major finding of this study is that the secretory cells of Drosophila female reproductive glands are required for efficient ovulation. When secretory cell number is deficient initially, ovulation is drastically reduced, despite normal copulation and the presence of sperm in the reproductive tract. When secretory cell function is reduced during adulthood by knocking down expression of Hr39, the time required for ovulation is greatly increased. Unlike the secretions that attract and stabilize sperm, ovulation is not disrupted by knocking down the protein secretory pathway. However, there are many possible reproductive tract secretions that might be released from the gland secretory cells by other mechanisms.

A previous study by Schnakenberg et al. (Schnakenberg et al., 2011) examined the role of spermathecal secretory cells by partially ablating them during adulthood using secretory protein regulatory elements to drive the apoptosis inducer Hid. Like these authors, we found that reproductive tract secretions are required to attract sperm to the spermathecae and to maintain their normal structure. We extended these observations by showing that a minimum number of about 80 secretory cells is required for normal sperm attraction and storage, and that these functions require the canonical protein secretory pathway and Hr39 expression. Schnakenberg et al. reported that egg release from the uterus frequently but sporadically is reduced in females with a deficit of secretory cells, and suggested that secretory cells produce an initial, long-lasting lubricant that coats the uterus. In contrast, when secretion was compromised, we saw that egg laying was strongly and consistently reduced due to defects in ovulation rather than in egg release. Schnakenberg et al. did not study ovulation independently from egg laying. In contrast, we used assays that separate these processes, allowing the role of secretion in ovulation to emerge.

The control of ovulation may be conserved

In mammals, mature Graffian follicles compete for ovulation based on complex hormonal and biochemical signals that are closely tied to products locally produced by the follicle's granulosa and nascent luteal cells (Mihm and Evans, 2008). Much less is known about how individual follicles in the Drosophila ovary are selected for oviduct entry from a large pool. While nervous control of ovulation is clearly required to coordinate egg release with environmental and circadian factors (Yang et al., 2008), the underlying mechanism of egg selection is likely to be more complex and involve local interactions as well as communication between the ovaries. Identification of the secretory cell product(s) that are required for ovulation would provide an important clue to uncovering these mechanisms.

A particularly attractive possibility is that communication signals between the reproductive tract and the ovary have been partially conserved between mammals and Drosophila. This prospect is now strengthened by the finding that Hr39 is required for ovulation, like its mammalian counterpart, Lrh-1. Prostaglandin-like molecules are known to regulate ovulation in mammals (Dinchuk et al., 1995; Lim et al., 1997), and Lrh-1 is thought to function by regulating expression of the prostaglandin-generating enzyme COX-II in mouse granulosa cells (Duggavathi et al., 2008). A prostaglandin-like signal already is known to function during egg maturation in Drosophila (Tootle and Spradling, 2008), and a COX-II-like enzyme, CG10211, is expressed in spermathecae under the control of Hr39 (Allen and Spradling, 2008). It will be interesting to determine if Lrh-1 functions in reproductive tract secretory cells. Finally, identifying additional glandular products acting in the Drosophila reproductive tract may elucidate additional pathways of communication between oviduct and ovary that are relevant to the induction of ovarian cancer.

Materials and methods

Drosophila genetics

Flies were reared on standard cornmeal-molasses food at 25°C unless otherwise indicated. Trans-heterozygous combinations of Hr397154/Ly92 (Allen and Spradling, 2008) and lz3/34 (from Bloomington Drosophila Stock Center, BDSC, Bloomington, IN) were used for loss-of-function analysis. For the rescue experiment with fruGal4 driving UAS-mSP (Yang et al., 2009), Hr397154/Cyo; fruGal4 females were crossed to Hr397154/Cyo; UAS-mSP or Hr397154/Cyo. Knockdown with RNAi or dominant negative constructs were carried out at 29°C and the following lines were used: UAS-cycARNAi (V32421; Vienna Drosophila RNAi Center, Vienna, Austria), UAS-cdc2RNAi (V41838), UAS-Hr39RNAi (V37694), UAS-hntRNAi1 (V101325), UAS-hntRNAi2 (V3788), UAS-NRNAi (gift from Sarah Bray), UAS-PsnDN (UAS-Psn.527.D447A, BDSC), UAS-Su(H)DN (Mukherjee et al., 2011), UAS-mCD8:GFP, lzGal4 (BDSC), dpr5Gal (51B02; Pfeiffer et al., 2008), ppkGal4 (Yang et al., 2009). In order to inhibit membrane recycling, the canonical exocytosis pathway, or Hr39 function in adult secretory cells, syt12Gal4 (47E02; Pfeiffer et al., 2008) was crossed to UAS-shits (Yang et al., 2009), while UAS-dcr2; syt12Gal4, tubGal80ts was crossed to the RNAi line against bCOP (BDSC 33741), sec23 (BDSC 32365), or Hr39 (V37694) at 18°C. Virgin females were selected 4 hr after eclosion and immediately shifted to 29°C for 6 days. UAS-RFP was used to monitor syt12Gal4 expression. For lineage tracing experiments, specific Gal4 driver was crossed to G-Trace lines to monitor real-time expression and lineage expression (Evans et al., 2009). Clonal labeling and pupae preparation were as previously described (Sun and Spradling, 2012). The Notch activity reporter Su(H)GBE-Gal4, UAS-mCD8:GFP was used to monitor Notch activation (Zeng et al., 2010), and ProtB-GFP was used to visualize sperm DNA (Manier et al., 2010). Control flies were derived from specific Gal4 driver crossed to wild-type Oregon-R.

Egg laying, ovulation, and copulation tests

4- to 6-day-old virgin females were fed with wet yeast 1–2 days before egg laying experiments. Five females were mated to 10 Oregon-R males in each bottle covered with the molasses plate at 25°C and the number of eggs was counted every day for 2 days except for experiments that perturb the exocytosis pathway or Hr39 function in adult secretory cells, which were carried out at 29°C for 2 days. For ovulation and copulation tests, single-pair matings between a 4- to 6-day-old virgin female and a ProtB-GFP male were carried out in the morning at 25°C, except for experiments that perturb the exocytosis pathway or Hr39 function in adult secretory cells, which were carried out at 29°C. 6 hr after mating, females were dissected to examine eggs inside the reproductive tract, and the corresponding vials were also examined for laid eggs. Female reproductive tracts were then fixed with paraformadehyde and sperm inside them were examined to determine copulation success. The number of sperm inside spermathecae was manually counted. In Table 2, egg laying time (in minutes) = 22 × 60/number of eggs; the ovulation time = the egg laying time × (1 − egg distribution in uterus); and uterus time = the egg laying time × egg distribution in uterus. The 95% confidence intervals were calculated correspondingly.

Immunostaining and microscopy

Pupal and adult reproductive tract staining was carried out as previously described (Sun and Spradling, 2012). Briefly, tissues were dissected in Grace's media, fixed in 4% EM Grade Paraformadehyde for 15–20 min, and blocked in PBTG (PBS + 0.3% Triton + 0.5% BSA + 2% normal goat serum). Incubation with primary antibody overnight was followed by a 2-hr incubation with secondary antibody and DAPI staining. Tissues were then mounted in Vectashield mounting media. The following primary antibodies were used: mouse anti-Hnt (1:75; Developmental Study Hybridoma Bank), rabbit anti-GFP (1:4000; Invitrogen), and chicken anti-β–Gal (1:1000; Abcam). Secondary antibodies were Alexa 488 and 546 goat anti-mouse, anti-rabbit, and anti-chicken (1:1000; Invitrogen). Images were acquired using the Leica TCS SP5 confocal microscope or the Zeiss Axioimager ZI microscope, and assembled using photoshop software.


  1. 1
  2. 2
  3. 3
    A study of the factors affecting fertility of lozenge females of Drosophila melanogaster
    1. RC Anderson
    Genetics 30:280–296.
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
    Ecdysone response genes govern egg chamber development during mid-oogenesis in Drosophila
    1. M Buszczak
    2. MR Freeman
    3. JR Carlson
    4. M Bender
    5. L Cooley
    6. WA Segraves
    Development 126:4581–4589.
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
    Mated Drosophila melanogaster females require a seminal fluid protein, Acp36DE, to store sperm efficiently
    1. DM Neubaum
    2. MF Wolfner
    Genetics 153:845–857.
  43. 43
  44. 44
  45. 45
  46. 46
    Developmental genetics of oogenesis
    1. AC Spradling
    In: B Martinez-Arias, editors. The development of Drosophila melanogaster. Cold Spring Harbor: Cold Spring Harbor Laboratory Press. pp. 1–70.
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57

Decision letter

  1. Utpal Banerjee
    Reviewing Editor; University of California, Los Angeles, United States

eLife posts the editorial decision letter and author response on a selection of the published articles (subject to the approval of the authors). An edited version of the letter sent to the authors after peer review is shown, indicating the substantive concerns or comments; minor concerns are not usually shown. Reviewers have the opportunity to discuss the decision before the letter is sent (see review process). Similarly, the author response typically shows only responses to the major concerns raised by the reviewers.

[Editors' note: this article was originally rejected after peer review. After an appeal against the decision and discussions between the authors and editors, the editors agreed that the submission could be published, subject to revisions. Both decision letters follow, along with the authors' response to the request for revisions.]

Original decision letter including full reviews

Thank you for choosing to send your work entitled “Ovulation in Drosophila requires reproductive tract secretions controlled by the nuclear hormone receptor Hr39” for consideration at eLife. Your article has now been peer reviewed and we regret to inform you that your work will not be considered further for publication at this time. Your submission has been evaluated by 3 reviewers, one of whom is a member of our Board of Reviewing Editors, and the decision has been discussed further with one of eLife's Senior editors. The Reviewing editor and the outside reviewers discussed their comments before we reached this decision.

Each reviewer has acknowledged the importance of the work, but all three are unanimous in suggesting that the work presented is not complete and ready for publication. Reviewer 1 points to the lack of a detailed developmental analysis or identification of gene function, aspects that echo the comments of Reviewer 3. Similarly, Reviewer 2 feels that the nature of the secreted product, indeed the nature of secretion itself, is not fully examined, a view shared by Reviewer 1. Finally, Reviewer 3, an expert in the specific field, raises many technical questions and also would like to be further convinced that some of the points raised in different parts of the manuscript have indeed been settled by the results that are presented. The detailed comments of each reviewer are included below.

Reviewer #1:

The paper “Ovulation in Drosophila requires reproductive track secretions controlled by nuclear hormone receptor Hr39” explores an interesting link between secretions from female reproductive tract glands (Spermathecae and Parovaria) with ovulation in Drosophila. The authors show that a developmental defect in the specification of secretory organs in the female reproductive tract (through mutations in genes such as lozenge and Hr39) causes significant ovulation defects. These defects are neither due to mating deficiencies, nor caused by defects in neurons that innervate the oviduct muscles.

Based on the stereotypical division patterns seen in cells of the secretory unit precursor (SUP) cells and the expression of Notch signaling reporters at high levels, the authors test and establish that Notch signaling is required for the specification of secretory cells. Similarly, the authors establish a role for the transcription factor Hindsight, which is also expressed in these cells. Finally, using a Janelia Gal4 driver line expressed specifically in developing but not in mature secretory cells, the authors establish that secretory cells are needed in sufficient numbers to allow normal ovulation and sperm storage. Previous studies have shown that secretory cells in the female reproductive glands release their secretions into a structure called the end apparatus, which is in close apposition to the microvilli of the secretory cells. The authors test the role of secretions from these cells using a specific Gal4 and block exocytosis from these cells by over-expressing a dominant negative version of Dynamin (shi[ts]). This causes a defect in sperm storage but not in ovulation. The authors conclude that either the secretory factors from SC that regulate ovulation are independent of exocytosis or they are needed at very low levels, not disrupted in the chosen experimental design.

This is largely an observational paper, with few broader insights to report. That the secretory cells control ovulation, and that this process is Notch dependent is an interesting finding, described and proven with well-planned experiments and originality in the paper. But the manuscript does not explore much beyond this basic premise.

The data presented are of high standard but the paper lacks a cohesive narrative that moves the field forward. For example, the paper does not explore in detail the developmental aspects of the SC cells, and uses an assortment of mutations (in Hr39, lz, Hindsight, Notch) only as ways to eliminate the function of these cells rather than explore their interdependencies, or developmental relevance. Similarly, no attempt is made on the functional side to determine the nature of the secretions that have such profound effects on ovulation. In the end, the negative result with shibire-ts leaves the question wide open for further exploration. This work represents an exploratory project that has a lot of future potential but is not yet ready to move the field forward.

Reviewer #2:

The findings of Spradling and colleagues are potentially interesting. The paper clearly shows the requirement of secretory cells in the events leading to attraction and storage of the sperm. However, the data on the nature of the secretory components is not satisfactory. The authors report the involvement of Shibre (Dynamin); however, shibre is known to affect endocytosis in the neurons: the significance of this mutation on exocytosis is, therefore, unclear. It is important to know if the defect is due to secretion of extracellular matrix components or some growth factors or cytokines. I do not expect the authors to reveal the identity of the essential secreted component(s); they must, however, provide the clear involvement of the secretory process. Does it involve the regulated or the constitutive secretory pathway? The authors should test the effect of mutations in the genes that block constitutive (ER-Golgi-plasma membrane) and the regulated (Golgi-secretory storage granules-PM) in sperm attraction and attachment to support their proposal.

Reviewer #3:

There are a number of interesting conclusions in this paper; however, the title suggests a link between Hr39 and secretory cell function in the spermathecae, but the only clear role suggested by the experiments for Hr39 is a role in the formation of the spermathecae and the parovaria, not in the functioning of fully formed secretory cells in these tissues. In fact, using Hr39 RNAi with the dpr5-Gal4 driver showed what appears to be a small drop in ovulation after 6 hrs (from 64.5 to 45.8 although the RNAi control is missing) – however, no drop in egg laying after 48 hours was seen. In addition, no significant drop in sperm storage was seen. Based on these findings, it is difficult to see why the authors used Hr39 in the title of this paper. In addition it seems that the interesting conclusions are not highlighted in the Abstract or Introduction. There needs to be a clear differentiation in the writing of the paper between conclusions based on the absence of the entire tissue, and those effecting cell number on adult function. Overall, the paper should be rewritten to clearly highlight the more interesting conclusions regarding the role of the secretory cells of the spermathecae (i.e. ovulation defects and sperm storage defects), and the conclusions based on hr39 and lz should be more clearly stated for what they are, a study of the loss of the tissues they help form.

Specific comments:

* The authors should state in the introduction the previously established link between the secretory cells of the spermathecae and sperm storage/ovulation – although Schnakenberg et al., (2011) is mentioned in the Discussion, it should also be mentioned up front in the Introduction.

* The authors need to clearly state that lzGal4 expression has previously been shown to recapitulate actual lz expression. When does the lzGal4 line come on in development?

* The authors state: “lz is expressed in the subset of these neurons near the oviduct-uterus junction, but the number of fru+ppk+ neurons was not affected by knocking down cycA or hr39 using the lzGal4 driver.” How do the authors know that knocking down cycA or hr39 should change neuronal numbers? Do they know if lz is required for their proliferation? What if it has another function in these neurons?

* Why is hr39 knockdown using RNAi tested in lz cells, but not lz-RNAi itself? Has hr39 previously been shown to be expressed in these lz neurons? ppk-Gal4 with hr39-RNAi does not effect egg laying – why don't the authors test lz?

* The result in Figure 2E is not that mSP expression in fru+ neurons induces egg laying, an already published finding (Yang et al., 2009), but that mSP expression in fru+ neurons can not bypass the egg-laying defect in hr39-/- mutants. Therefore, the SP behavioral switch can happen in the absence of Hr39 (and therefore no spermathecae or parovaria are needed); however, the loss of egg-laying in hr39-/- cannot be overcome by the SP behavioral switch.

* It appears from the methods that the shi[ts] flies are shifted just prior to mixing males and females. Is one possible explanation for the shi[ts] phenotype that all the important secretions for ovulation/egg-laying are already in place prior to mating?

* There appears to be no mention of statistical analysis: the types of analysis should be clearly stated along with every figure/table where they have been used and statistically significant differences/similarities should always be highlighted.

* fru+/ppk+ neurons have recently be shown to also be dsx-positive (Rezával et al., 2012).

* There are many different Gal4 lines used in this study: the authors should make it clear throughout the paper exactly where the lines have been shown to be expressed (either previously published or by the authors themselves). For example, the syt12-Gal4 line is introduced near the end of the paper but its overall expression pattern and reasons for use are far from clear.

* The first paragraph of the Discussion is confusing (for example, using double-negatives): this section should be re-worked to clearly state what was previously shown, and then clearly state what the current study adds to the story. The authors have shown a clear ovulation defect that was previously not shown; this is an interesting result that currently gets lost in the detailed explanation of the previous study.

Decision letter outlining revision requirements

Thank you for choosing to send your work entitled “Ovulation in Drosophila requires reproductive tract secretions controlled by the nuclear hormone receptor Hr39” for consideration at eLife, and for telephonic discussions with Detlef Weigel and K. VijayRaghavan on December 21, 2012, which were useful. This has been followed up with discussions amongst the editors and we suggest the following points for a revision of the manuscript.

1)We strongly encourage you to change the title because it has been a source of confusion about what the paper is about. The current title can easily be read as implying that the paper will reveal the identity of the Hr39-dependent secretions. The title should instead reflect that this paper shows for the first time an unexpected way by which ovulation is controlled by the female reproductive tract.

2)The clear responses given to Referee 1 could be included in the text so that the paper better documents the assertion that this is indeed a foundational paper.

3)Next, experiments documenting the adult role of Hr39 should be included. Experiments that report the effects of disrupting at least one key gene in the ER/Golgi/PM secretory pathway(s) on ovulation would also greatly strengthen the manuscript. Your appeal letter indicates that both types of experiments are readily feasible.

4)Finally, a measured and brief analysis of your views on the Schnakenburg et al. paper, in the context of the current work, will be useful for readers to compare the two studies and to reach their own conclusions about the similarities and differences between them.


Author response

Following the previous review, the editors made four requests for a revised version of our manuscript:

1) We strongly encourage you to change the title because it has been a source of confusion about what the paper is about. The current title can easily be read as implying that the paper will reveal the identity of the Hr39-dependent secretions. The title should instead reflect that this paper shows for the first time an unexpected way by which ovulation is controlled by the female reproductive tract.

We changed the title to “Ovulation in Drosophila is controlled by secretory cells of the female reproductive tract” to avoid any implications that specific secretions have been analyzed. “Hr39” was removed from the title as requested.

2) The clear responses given to Referee 1 could be included in the text so that the paper better documents the assertion that this is indeed a foundational paper.

We extensively rewrote the Abstract, changed a few lines in the Introduction, and modified the paragraph order in the Discussion to highlight the importance of the paper for studies of ovulation. A new section of Results and portions of the Discussion were modified to accommodate the new experiments. To further address why ovulation rather than Notch signaling should be a focus (as requested), we added a brief note in the section on Notch signaling.

3) Next, experiments documenting the adult role of Hr39 should be included. Experiments that report the effects of disrupting at least one key gene in the ER/Golgi/PM secretory pathway(s) on ovulation would also greatly strengthen the manuscript. Your appeal letter indicates that both types of experiments are readily feasible.

The main reason for the delay in resubmitting the revised version was to carry out both of these experiments. To unequivocally test the adult role of Hr39 (and the ER/Golgi/PM pathway) in adults, we drove RNAi expression targeting Hr39, betaCOP or Sec23 with Syt12-Gal4 in the presence of GAL80ts. At 18 °C the GAL80 inhibitor prevents Gal4 expression and the flies are phenotypically wild type with a normal number of secretory cells. We raised the temperature to 29°C at the time of adult eclosion. Some secretory products have probably been produced already at this time, and the turnover of the GAL80 protein at the restrictive temperature can take a few days, providing additional time for additional normal secretory cell function. Despite the potential for perdurance, when we tested the ability of these flies to store sperm, ovulate, or lay eggs at day 6 of adulthood, we saw strong defects (described below), thereby proving that Hr39 gene function and secretory cell function are required in adulthood. These results are presented in a revised Figure 5 and a new table: Table 2. The first two bars of panels D, G, and H present the original results, while the remaining bars show the new experiments. In addition, we added a new panel to Figure 5 with a timeline for the temperature shift and assay to make it clear that we are testing only adult function.

With respect to sperm storage, knocking down Hr39 in adulthood caused an extensive and statistically significant reduction in the ability of females to store sperm following mating (Figure 5D). When either betaCOP or sec23 was knocked down we observed even stronger effects on sperm storage than previously documented with shi[ts], in that all spermathecae now contain 15 or fewer stored sperm (see new Figure 5D). The few sperm that were present showed the same defective morphology illustrated in Figure 5F for shi[ts]. Moreover, knocking down betaCOP or sec23 by shifting the temperature beginning in 9-day-old adults (rather than at eclosion) and testing at 14 days of age gave the same result.

When we quantitatively analyzed egg laying and ovulation, these experiments also confirmed and extended our previous studies. We previously showed that shi[ts] did not cause a significant reduction in egg laying (over 24 hr) or in ovulation rate (measured in the first 6 hours after mating). We improved on these assays by measuring both the egg laying rate of the flies over 2 days beginning at day 6 (Table 2, column 1) and for a 6 hour sample of this period, measuring in individual flies of each genotype the number of eggs laid and the presence or absence of an egg still in the uterus (Table 2, column 2). (We confirmed that the percentage of flies with uterine eggs was the same after 6 hours and after the longer 2-day period.) By calculating the total time required for each laid egg, and by using the percentage of uterine eggs, we apportioned that time between ovulation time and time in the uterus (Table 2, column 3). Ovulation time provides a sensitive measurement of how efficiently ovulation is occurring in the different genotypes.

These results show that egg laying is not significantly reduced over 2 days by blocking secretion from reproductive secretory cells using shi[ts], betaCOP RNAi or sec23RNAi (Figure 5F), although it does trend downward, most likely due to secondary effects. Strikingly, ovulation time is not increased at all by disrupting ER/Golgi/PM secretion in these cells, confirming our previous result with shi[ts]. In contrast, Hr39-RNAi in secretory cells does significantly decrease the number of eggs laid and substantially and significantly slows ovulation (Fig. 5H). This confirms our conclusion that sperm storage requires ER/Golgi/PM secretion from female secretory cells and shows that this requirement exists in adulthood. Moreover, these data now show clearly that female secretory cells also function in adulthood to stimulate ovulation by a process requiring adult Hr39 expression, but that the stimulation of ovulation by secretory cells is independent of the canonical secretory pathway. We suspect that this second secretion is likely to be a small molecule that can move more easily than a large protein from the spermathecae and/or parovaria to the base of the ovary.

4) Finally, a measured and brief analysis of your views on the Schnakenburg et al. paper, in the context of the current work, will be useful for readers to compare the two studies and to reach their own conclusions about the similarities and differences between them.

We have added such comments to the Discussion as a second paragraph in the initial section on ovulation. Rather than discussing all the issues raised by their paper, we summarized the areas of agreement and the major difference: Schnakenburg et al. were unable to analyze ovulation whereas we discovered this connection between reproductive tract secretion, Hr39 expression, and ovulation.

Additional changes:

The Materials and methods and Figure legends have been revised to describe the new experiments. In addition, error bars that had been inadvertently omitted in the original version were restored, and additional statistical test information was added to Figure legends or the Materials and methods.


Article and author information

Author details

  1. Jianjun Sun

    1. Department of Embryology, Carnegie Institution for Science, Baltimore, United States
    2. Howard Hughes Medical Institute, Baltimore, United States
    JS, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article
    Competing interests
    The authors declare that no competing interests exist.
  2. Allan C Spradling

    1. Department of Embryology, Carnegie Institution for Science, Baltimore, United States
    2. Howard Hughes Medical Institute, Baltimore, United States
    ACS, Conception and design, Analysis and interpretation of data, Drafting or revising the article
    For correspondence
    Competing interests
    The authors declare that no competing interests exist.


Howard Hughes Medical Institute

  • Allan C Spradling

The funder had no role in study design, data collection and interpretation, or the decision to submit the work for publication.


We are grateful to Dr Gerald Rubin for allowing us screen through the Janelia Gal4 collection. We also thank Drs Yuh Nung Jan, Sarah Bray, Utpal Banerjee, Scott Pitnick, and Steve Hou for sending us fly lines; Drs Chen-Ming Fan, Ming-Chia Lee, Vicki Losick, Matt Sieber, and Ethan Greenblatt for comments and discussion on the manuscript. ACS is an Investigator of the Howard Hughes Medical Institute.

Reviewing Editor

  1. Utpal Banerjee, Reviewing Editor, University of California, Los Angeles, United States

Publication history

  1. Received: November 21, 2012
  2. Accepted: March 8, 2013
  3. Version of Record published: April 16, 2013 (version 1)


© 2013, Sun and Spradling

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.


  • 1,610
    Page views
  • 209
  • 25

Article citation count generated by polling the highest count across the following sources: Scopus, Crossref, PubMed Central.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

  1. Further reading

Further reading

    1. Cell Biology
    2. Plant Biology
    Amutha Sampath Kumar et al.
    Research Article Updated